|
32 posts total
12
 For the Verge, @sarahjeong reached out to me to see if I’d be open to writing about my personal experience being colorblind and issues around accessibility. It’s currently the top story on their homepage, and they did a beautiful job with it. I hope you like it. https://www.theverge.com/23650428/colorblindness-design-ui-accessibility-wordle Found a weird thing hiding in macOS! https://waxy.org/2023/04/the-bitcoin-whitepaper-is-hidden-in-every-modern-copy-of-macos/
Show previous comments
@andybaio Whatever this story is about (what? Apple hired a dickhead? Heaven forfend), it’s the diametric opposite of digging up a monolith on the moon.  Just found this browsing Twitch’s “AI” tag: a 24/7 livestream of deepfaked Twitch streamers and obnoxious personalities (e.g. Elon Musk, Jordan Peterson, Joe Rogan, Ben Shapiro) answering questions from chat. It’s very uncanny valley: avatars move around, drink water, etc. https://www.twitch.tv/atheneaiheroes Material for the deepfaked video/audio/dialogue is all crowdsourced in their Discord: video clips, voice clips, catch phrases, details on their personal opinions and biography… Waiting for someone to make an artisanal search engine that only indexes manually verified websites written by humans. Farm-to-table organic search results, handmade with 100% certified human creativity, like your grandparents told you about. Judging from all the “that’s Yahoo” replies, it seems like many people have forgotten the difference between a directory and a search engine? A directory is an organized list of websites (e.g. Yahoo, DMOZ, ooh.directory), while a search engine usually searches the text of web pages (e.g. Google). I want a *search engine* that crawls and searches pages from a curated list of websites (which could be a directory). Marginalia is the closest I’ve seen. https://search.marginalia.nu/ Google used to take pride in minimizing time we spent there, guiding us to relevant pages as quickly as possible. Over time, they tried to answer everything themselves: longer snippets, inline FAQs, search results full of knowledge panels. Today's Bard announcement feels like their natural evolution: extracting all value out of the internet for themselves, burying pages at the bottom of each GPT-generated essay like footnotes. https://blog.google/technology/ai/bard-google-ai-search-updates/ Of course, they frame it as "making it easier for people to get to the heart of what they’re looking for and get things done." But how long will publishers, blogs, online communities, and other creators tolerate generating free content to feed their machine, while getting little to nothing in return? It's only a one line change: Mastodon Flock is an excellent Twitter-to-Mastodon friend finder: all the functionality of Movetodon (before Twitter killed it today), but open-source and with a delightfully retro interface. https://mastodon-flock.vercel.app/   Twitter just announced free access to the Twitter API will end on February 9, with unannounced pricing or usage details, effectively killing every free fun and useful thing ever built with it: bots, games, mashups, visualizations, research projects, autoposters, autoblockers, deleters, and so much more. https://twitter.com/twitterdev/status/1621026986784337922 I guess that means the end of my own Belong.io link aggregator, which I’ve run for nearly ten years. https://belong.io/ I really loved this long interview with Matt Mullenweg about the Tumblr acquisition. Nilay Patel covers a lot of ground about content moderation, staffing challenges, and how to responsibly acquire a social network: finding a path to sustainability without alienating its community and killing what makes it special. https://www.theverge.com/23506085/wordpress-twitter-tumblr-ceo-matt-mullenweg-elon-musk  @andybaio "I feel like when you’re hosting a social network, it’s your responsibility to provide a safe and healthy environment." Good thinking! if anyone needs a four-foot-tall Twitter logo or a six-foot-tall @ symbol, twitter's having a fire sale https://www.bidspotter.com/en-us/auction-catalogues/heritage-global-partners/catalogue-id-herita10194 maybe the auction is trying to earn a few bucks since these are the only kinds of advertisers left Once again, @liza turned in a stunning entry for NaNoGenMo, an annual project to write code that generates a novel-length work. This year's entry is A Letter Groove, cutting out words from books, revealing the pages underneath. https://post.lurk.org/@liza/109297612410850588 I've written about her NaNoGenMo entries in the past: generative blackout poetry and Seraphs, a procedurally-generated codex inspired by the Voynich Manuscript. https://waxy.org/2016/11/liza-dalys-generative-blackout-poetry/ I talked to the The Verge for this deep dive into the muddy waters of generative AI and copyright. It’s complicated! https://www.theverge.com/23444685/generative-ai-copyright-infringement-legal-fair-use-training-data  @andybaio ... fundamentally, don't artists find inspiration in other artists? In doing so, do those styles inadvertently influence the art created? Could traditional art be considered generative (to a certain degree)? Last week, a Redditor fine-tuned an AI image model on the work of one illustrator, sparking a debate about the ethics of reproducing a living artist's style. I talked to that artist to see how she felt about it, and the person who made it. https://waxy.org/2022/11/invasive-diffusion-how-one-unwilling-illustrator-found-herself-turned-into-an-ai-model/ The Redditor used a technique developed by Google called DreamBooth, reimplemented for Stable Diffusion, using only 32 illustrations from the artist. Training took 2.5 hours of cloud GPU time at a cost of under $2. He then released the finetuned model for everyone to use. 12
|
Gregory's Server